Turns Out the Internet Isn't Great at Preserving Scientific Research

There was a period in the early days of the Internet becoming a part of everyday life when it seemed like it might become a permanent online home for, well, everything. If you’re old enough, you might remember the phrases “long tail” and “celestial jukebox” used admiringly — though even in the first decade of this century, there was some sense that this might not all go according to plan.

Now it’s 2024 and we have a better sense of what the Internet is and is not good for and, unfortunately, if you’d been banking on the idea of it as a permanent repository for every research paper ever written, brace yourself for what’s coming next. Writing at Nature, Sarah Wild has news of a recently published paper that found that over 2,000,000 studies stored online appeared to have vanished.

This study of studies, published in the Journal of Librarianship and Scholarly Communication, focused on 7.4 million works that had been given digital object notifiers. The good news first: well over half of the DOIs in the study could be located in one archive or more. The bad news? 27.64% of the DOIs cited were “seemingly unpreserved.”

“If you can’t verify what someone else has said at some other point, you’re just trusting to blind faith for artifacts that you can no longer read yourself,” the study’s author, Martin Eve, told Nature.

What Can We Learn From Jack Kerouac’s Archives?

The writer would have been 100 this week

It’s a broader issue than the discourse over what happened to the Long Tail, but the two seem inherently connected. At the very least, both reckon with questions of digital preservation — and the frustrations of what can happen if and when something simply vanishes from the digital realm. That could be a research paper or a favorite recording — but either way, it raises big questions about how we preserve knowledge.